Mainstreet has released the results of several riding-level polls in the Ottawa area, giving us better data on how each party is doing than we’ve had to date. Until now, in the absence of such poll results, I’ve been making do with ThreeHundredEight.com’s riding-by-riding projections. These projections should be taken with a grain of salt: their methodology involves applying a multiplier based on current regional polls to the results the last time around (e.g., if a party is doing 10 points better in the polls than they did in the last election, apply that multiplier to the results in each riding).
One of the five ridings Mainstreet polled was mine, Pontiac. So now not only do we have a better sense of what’s going on in this riding, we also have a cautionary tale of just how much salt we need to take those riding-by-riding projections with. ThreeHundredEight.com has for the past week or two been characterizing the Pontiac as a close three-way contest, with the Conservative candidate, Ben Woodman, holding a slim lead until today; the Liberal candidate, Will Amos, is now projected to have a 50 percent chance of taking the contest.
Except that the Mainstreet poll says it’s not a three-way contest. They have Woodman well behind, with only 17 percent of the vote. According to Mainstreet, it’s a two-way contest between Amos and the incumbent NDP MP, Mathieu Ravignat, with Ravignat leading Amos 29 to 28 percent. Among decided and leaning voters, Ravignat leads by two points, 36 percent to Amos’s 34 percent, with Woodman bringing up the rear at 19 percent.
(Note that in 2011 the Pontiac constituency had substantially different boundaries.)
The poll had a sample size of 673. Its margin of error is ±3.74 percent, which is to say that Amos and Ravignat are essentially tied. But the story doesn’t necessarily end there, because it matters how the Pontiac riding was sampled. As I’ve shown elsewhere, different parts of the Pontiac vote in different ways: last time around the Conservative candidate won the Pontiac MRC proper, split the Gatineau Valley with the NDP and got roundly clobbered in Gatineau proper and the surrounding exurbs. If Mainstreet oversampled the urban and exurban areas, Woodman might actually be doing better than the poll suggests — but not necessarily by much, because the deep rural areas only make up a third of the riding’s population. And vice versa, if they oversampled the rural areas.
If nothing else, these poll results may give some relief to anyone who was planning on voting strategically and agonizing over the choice: you can vote for whomever you want without fear of splitting the vote. (Mind you, I think that in a close three-way race strategic voting is pointless, because it’s impossible to know what the right call is — but that’s a subject for another post.)